Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, reaching epidemic proportions in the aging population and causing considerable morbidity, mortality, and socioeconomic burden. AF is primarily a disease of older people-84% are older than 65 years. Moreover, the incidence of AF sequelae such as stroke increases with advancing age. Recently, indices of arterial structure and function, such as carotid intima-media thickness (cIMT) and pulse pressure (a surrogate of arterial stiffness) have been shown to be associated with AF. There is a potential mechanism for this association: structural and functional changes in the vasculature lead to hypertension, left ventricular hypertrophy, and finally, structural remodeling in the atria. However, several knowledge gaps remain. It is unknown whether cIMT measurement improves risk prediction of AF, over and above established risk factors for AF. In addition, other indices of arterial stiffness such as carotid distensibility have not been studied in relation to AF in prospective cohort studies. Furthermore, very little is known about whether cIMT, carotid distensibility, or pulse wave velocity (PWV) predicts important clinical sequelae of AF, such as ischemic stroke and death. Identification of other indices of arterial structure and function that improve prediction of AF an its sequelae will enable us to identify individuals who will benefit from preventive interventions, thus reducing AF burden in the elderly population. In addition, if cIMT, carotid distensibility, an PWV measurements are found to improve risk prediction of AF, these findings will reinforce an emerging conceptual viewpoint on the pathophysiology of AF, specifically, that AF is not purely an electrophysiological disorder, but may be a consequence of vascular disease. In this project, we will evaluate the extent to which cIMT, carotid distensibility, and PWV are associated with AF incidence. We will also evaluate the discriminatory power of these novel risk factors, the extent to which these risk factors improve accuracy and calibration of predictive risk models, and the ability of these risk factors to improve risk classification. Our overall hypothesis is that indice of arterial structure and function improve risk prediction of AF and its sequelae. We will test our hypothesis for cIMT and carotid distensibility in an NIH- sponsored prospective biracial cohort study-Atherosclerosis Risk in Communities (ARIC) and replicate our findings in the Cardiovascular Health Study (CHS) and the Rotterdam Study (RS). We will evaluate the role of PWV in the RS. We will leverage the extensive phenotypic data, more than 3,900 incident AF cases, more than 2,500 deaths and 500 incident ischemic stroke events in participants with AF in the 3 cohorts to achieve the following specific aims: 1) Identify arterial indices that improve risk prediction of AF, 2) Identify arterial indices that improve risk prediction of ischemic strokein AF, and 3) Establish which arterial indices enhance risk prediction of all-cause and cardiovascular mortality in AF.